EP1122927A2 - Route lookup engine - Google Patents

Route lookup engine Download PDF

Info

Publication number
EP1122927A2
EP1122927A2 EP00310758A EP00310758A EP1122927A2 EP 1122927 A2 EP1122927 A2 EP 1122927A2 EP 00310758 A EP00310758 A EP 00310758A EP 00310758 A EP00310758 A EP 00310758A EP 1122927 A2 EP1122927 A2 EP 1122927A2
Authority
EP
European Patent Office
Prior art keywords
lookup
route
block data
engine
bit
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
EP00310758A
Other languages
German (de)
French (fr)
Other versions
EP1122927A3 (en
Inventor
Andrew T. Hebb
Sanjay G. Cherian
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ascend Communications Inc
Original Assignee
Ascend Communications Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ascend Communications Inc filed Critical Ascend Communications Inc
Publication of EP1122927A2 publication Critical patent/EP1122927A2/en
Publication of EP1122927A3 publication Critical patent/EP1122927A3/en
Ceased legal-status Critical Current

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/901Indexing; Data structures therefor; Storage structures
    • G06F16/9017Indexing; Data structures therefor; Storage structures using directory or table look-up
    • G06F16/902Indexing; Data structures therefor; Storage structures using directory or table look-up using more than one table in sequence, i.e. systems with three or more layers
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/54Organization of routing tables
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q3/00Selecting arrangements
    • H04Q3/64Distributing or queueing
    • H04Q3/66Traffic distributors
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13097Numbering, addressing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13103Memory
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13106Microprocessor, CPU
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13141Hunting for free outlet, circuit or channel
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13204Protocols
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13216Code signals, frame structure
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/1329Asynchronous transfer mode, ATM
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13299Bus
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/1332Logic circuits
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13367Hierarchical multiplexing, add-drop multiplexing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13384Inter-PBX traffic, PBX networks, e.g. corporate networks
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13389LAN, internet
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04QSELECTING
    • H04Q2213/00Indexing scheme relating to selecting arrangements in general and for multiplex systems
    • H04Q2213/13399Virtual channel/circuits

Definitions

  • Network devices such as routers must perform route lookups as part of their processing operations.
  • a route lookup is necessary for every packet received by a router in order to determine where the packet is to be sent next
  • the received packets include a 32-bit wide destination address.
  • the destination address identifies a unique address to which the packet is to be forwarded.
  • a route lookup utilizes the destination address of the received packet in order to identify the next-hop address and exit port in a routing table. Given that the destination address is 32 bits wide, the number of possible destinations is greater than one billion.
  • An IP address comprises a network number and a host number.
  • the network number is of variable length and identifies a unique network. Concatenated to the network number is a host number.
  • the host number identifies a unique host system within a particular network.
  • a received network number part of the address is matched with multiple entries in a routing table to identify the best match.
  • the best match is the one with the largest matching number of bits in the first part in the address.
  • Known methods used to perform this matching are complicated and can take a long time. It would be desirable to provide a route lookup engine which can perform a route lookup quickly and efficiently.
  • a Route Lookup Engine for determining a next hop index.
  • the RLE receives a lookup key and performs a multi-bit trie search with prefix expansion and capture of a variable stride trie.
  • the data that the RLE returns comprises the next hop information and status flags.
  • the RLE uses a compact, field reusable data structure.
  • the RLE performs both unicast and multicast IP address lookups on Virtual Private Networks.
  • the RLE uses separate indexing and forwarding memories. The upper bound of the search time for the RLE is fixed regardless of the route table size.
  • a Route Lookup Engine is disclosed.
  • the RLE may be used in a variety of applications.
  • the RLE is part of an Internet Protocol (IP) routing subsystem for an Asynchronous Transfer Mode (ATM) switch.
  • IP Internet Protocol
  • ATM Asynchronous Transfer Mode
  • the routing subsystem of the ATM switch provides a fast and efficient route lookup mechanism.
  • Fig. 1 a top-level block diagram of a an ATM switch is shown.
  • the ATM switch includes a Input/Output Adapter (IOA) 52.
  • the IOA includes interface optics, Synchronous Optical Network (SONET) framers and High Level Data Link Control (HDLC) controllers.
  • the ATM switch also includes a Input/Output Module (IOM) 10.
  • the IOM includes a plurality of forwarding engines 20.
  • the IOM further includes a timing control unit 30, a DC power conversion and distribution unit 40, a switch fabric interface 50 and control processor 60.
  • the timing control unit 30 of the IOM provides system reference for local timing to the physical layer.
  • the Power Conversion and Distribution unit 40 of the IOM comprises a DC-DC converter. The -48 volts available from the backplane is converted to the required voltages for the module.
  • the switch fabric interface 50 connects each forwarding engine with other cards in the switch via the switch fabric. Data is carried in the form of ATM cells. Switch fabric congestion information is also carried back from the fabric and reformatted/summarized for the forwarding engine.
  • the control processor 60 comprises a Pentium microprocessor as a background general-purpose processor.
  • the forwarding engines 20 will be described in detail below.
  • the forwarding engine 20 comprises an inbound Segmentation and Reassembly (SAR) device 150 with associated control memory 160 and packet memory 170, an outbound SAR 180 with associated control memory 190 and packet memory 200, a Packet Header Distributor (PHD) 210, a Packet Classification Engine (PCE) 110, a Route Lookup Engine (RLE) 100, a forwarding processor 130, a cache 140 associated with the forwarding processor and a DRAM/SRAM 120.
  • SAR Segmentation and Reassembly
  • PHD Packet Header Distributor
  • PCE Packet Classification Engine
  • RLE Route Lookup Engine
  • the forwarding engine 20 processes inbound and outbound packets. For inbound packets, a packet is transferred to the inbound SAR 150, where the packet is stored in the packet memory 170. The inbound SAR 150 sends the highest Quality of Service (QoS) packet to the PHD 210.
  • QoS Quality of Service
  • the forwarding processor 130 checks status of the PHD and determines when inbound packets are ready to be received. The forwarding processor 130 retrieves the highest priority header from PHD FIFOs and stores them in their cache 140.
  • the forwarding processor 130 processes the header and sends required header elements simultaneously to both the RLE 100 and the PCE 110 where they are stored in egress queues.
  • the RLE 100 performs a public or VPN unicast lookup and then a subsequent multicast lookup if required, while the PCE 110 runs filter checks against the header elements.
  • the RLE 100 and PCE 110 pass status to the PHD 210 and store results in an egress queue which is polled by the forwarding processor 130.
  • the operation of the RLE 100 will be described in detail below.
  • the forwarding processor 130 polls and subsequently obtains the route lookup and classification results from which a new header is formulated. This new header is then written to the PHD 210 along with transmit information.
  • the inbound SAR 150 retrieves the newly formatted header, returns it to packet memory 170 with the rest of the packet, and schedules the packet to be segmented and transmitted to the switch fabric.
  • the forwarding engine 20 processes outbound packets as follows. ATM cells are transferred from the switch packet to the outbound SAR 180 where the entire packet is reassembled and stored in packet memory 200.
  • the outbound SAR 180 sends the header of the highest QoS packet to the PHD 210.
  • the forwarding processor 130 checks the status of the PHD 210 and determines outbound packets are ready to be received.
  • the forwarding processor 130 retrieves the highest priority header from the PHD FIFOs and stores them in their cache.
  • the forwarding processor 130 processes the header and may send required header elements simultaneously to both the RLE 100 and the PCE 110 where they are stored in egress queues.
  • the RLE 100 performs a public or VPN unicast lookup and then a subsequent multicast lookup if required, while the PCE 110 runs filter checks against the header elements.
  • the RLE 100 and PCE 110 pass status to the PHD 210, store results in an egress queue which is polled by the forwarding processor 130.
  • the forwarding processor 130 polls and subsequently obtains the route lookup and classification results from which a new header is formulated. This new header is then written to the PHD 210 along with transmit information.
  • the outbound SAR 180 retrieves the newly formatted header, returns it to packet memory 200 with the rest of the packet, and schedules the packet to be transmitted to the physical frame interface.
  • the RLE is a hardware search engine for determining the next-hop index for the forwarding processor 130.
  • the RLE comprises a forwarding processor (FP) interface 65, input and output FIFOs 40 and 50, a lookup controller 20, a lookup table DRAM 30, and a control processor interface 70.
  • the FP interface 65 is used to load the key in the search engine.
  • the key comprises the address of the root of the tree/the VPN ID unique to the physical port/logical port, the IP destination address, the IP source address, the order of the first level of the tree, and an optional request ID to be returned with the header for sequence error checking.
  • the control processor interface 70 is used for performing table updates between lookups.
  • the FP writes one or more cache lines to a common address for both the RLE and the PCE simultaneously.
  • the FP interface 65 is used to retrieve the search results from both the RLE and PCE simultaneously. They report together from the same address, with a portion of the result provided by the PCE and a portion of the result provided by the RLE.
  • the RLE includes FIFO or other interfaces to enable it to maximize throughput.
  • the input and output FIFOs allow several lookup requests from the FP to be outstanding at a given time. If those packets have different lookup times, the FIFOs smooth the lookup rate, by allowing several header bursts in and out of the RLE.
  • the RLE performs both unicast and multicast IP address lookup on Virtual Private Networks (VPNs). Each VPN is used as a pointer to the root of a table specific to that VPN. Multiple VPNs may map to the same route table via software indirection.
  • VPNs Virtual Private Networks
  • the Power Processor Controller submits a route lookup request to the PHD.
  • the hardware VST tables of the RLE are used to resolve the request with the resulting indices included with the response from the RLE. Once the indices are made available to the Power Processor Controller, they are used to directly access the forwarding information from the forwarding table.
  • the results from the route lookup comprise the next-hop index, valid lookup/no route found bits, status flags, a multicast lookup preformed bit, statistics information and an optional request ID for sequence error checking.
  • the RLE search is first performed on the IP destination address and subsequently on the IP source address if the packet is a multicast packet.
  • the search utilizes a multi-bit tree search with prefix expansion and capture.
  • the search terminates when a next-hop index is found or the end of the key is reached.
  • Hardware error checking terminates the search and reports an error if an attempt is made to continue past 64 bits of IP header search at the end of the IP source address.
  • An RLE manager of the routing subsystem manages the RLE memory.
  • the RLE memory is used to store the hardware Variable Stride Trie (VST) route tables for each of the configured Virtual Private Networks (VPNs) including the default VPN, VPNO.
  • VST Variable Stride Trie
  • Each hardware VST is guaranteed a minimum amount of contiguous memory space. The remaining memory space is made available as free pools of memory blocks. As routes are added to and deleted from the VST tables, memory blocks are allocated from and released to these memory pools.
  • the routing subsystem manages the memory
  • the routing subsystem ensures that the forwarding database resident on the memory of the Power Processor Controller remains in synchronization with the hardware VST tables resident on the memory of the RLE. Similarly, the software VST tables resident on the memory of the Kernel processor must remain synchronized with the hardware VST tables of the RLE.
  • the RLE includes Variable Stride Trie (VST) based route lookup tables which reside on a local memory. These VST tables correspond to one specific route. Each entry of the hardware VST table contains an index to the corresponding forwarding entry within a forwarding table resident on the Power Processor Controller memory.
  • VST Variable Stride Trie
  • VST route tables are built with a Variable Stride Trie data structure which supports best matching prefixes at a high rate.
  • the VST tables are built by partitioning the search key into discrete bit strings.
  • the search key is the IP address.
  • the number of VST levels specifies the maximum number of memory accesses during the route lookup. Thus, the time to complete the lookup is bounded but not deterministic.
  • the length of each VST level is known as the stride length and determines the number of elements within the node representing that partition.
  • the hardware route search data structure comprises nodes, each of which is an array of elements.
  • the number of elements in a node depends on the node order, where order is the number of bits of the search key (destination IP address) used to evaluate this node.
  • Each element has a type indicator that designates the NEXT node as either a leaf or the root of another node.
  • the order indicates the order of the NEXT node if the type of the current node is "node" and the prefix length of the leaf if the type of the current node is "leaf'.
  • the pointer field depends on the element type.
  • the pointer is an index into the foreground forwarding table and identifies the forwarding entry for the destination address/mask pair. If the type is "node", the pointer is a pointer into search space identifying the address of the next node to be evaluated.
  • the RLE table entry comprises: Bits 31 30 29 28 24 23 0 Field P (I) T (1) I (1) Order (5) RLE-Root/Flag.NHP where:
  • the RLE Root is used to start the search.
  • the RLE root contains the root pointer and the order of the first level of the tree.
  • the first lookup address is comprised of the root pointer plus an index
  • the index is derived from the upper bits of the destination address. The number of bits is variable and is indicated by the order of the first level. This address is used to lookup an element from the table. If this element is a leaf then the search is complete and the element is passed back to the PCE/RLE manager. If the element is a node, then the pointer is used for the next lookup.
  • the address for the next lookup is comprised of the above pointer plus an index.
  • the index is derived from the "next" bits from the destination address.
  • the exact bits are derived from the order of the last level, plus the order of this level. Therefore, the order of each level needs to be tracked in order to know the relevant bits of the Destination Address to be used for the index.
  • This address is again used to lookup an element for the table. If this element is a leaf, then the search is complete and the element is passed back to the PCE/RLE manager. If the element is a node, then the pointer is used for the next lookup. This is repeated until a leaf is found. If the number of accesses or levels exceeds a maximum number an error is returned instead.
  • the indirect bit When the indirect bit is set, the leaf flag is ignored, the order field is not used, and the current 24 bit pointer is used as a direct table index, with the route results returned from that indexed location.
  • the indirect function can also be invoked by setting the depth to zero.
  • a flowchart of the presently disclosed method 300 of performing a route lookup is shown.
  • the lookup is started upon receipt of a request at step 310.
  • a determination is made as to whether this is the first lookup for this request. If this is not the first lookup for the request than the previous values for the TYPE, INDIRECT, ORDER and RLE ROOT are used, as shown in step 330. If this is the first lookup for this request, then the values for the TYPE, INDIRECT, ORDER and RLE ROOT are taken from the request block data.
  • the TYPE bit is evaluated. IF the type is not a "leaf', then steps 420 through 480 et seq. are executed. At step 420 the dram address is set to the RLE ROOT plus zero concatenated with the destination address and the source address.
  • Step 430 is then executed which performs a read from the dram address and new values for the TYPE, INDIRECT, ORDER and RLE ROOT are obtained.
  • the current order is set to the current order plus the order.
  • step 450 a determination is made as to whether the current order is less than or equal to sixty-four. If it is not, then an error condition is indicated and the search terminates. If the current order is less than or equal to sixty-four, then a determination is made.at step 470 as to whether the current order is greater than thirty-two: If not, then another lookup is performed beginning at step 320. If the current order is greater than thirty-two, then step 480 is executed the L bit is set, indicating a multicast operation. After step 480, another lookup is performed beginning at step 320.
  • step 360 a determination is made at step 360 whether the INDIRECT bit is set. If the INDIRECT bit is not set, then the lookup is complete and the results stored, as shown in step 410.
  • step 370 is executed wherein the dram address is set to the RLE ROOT.
  • step 380 the address of the RLE dram is read and new values for the TYPE, INDIRECT, ORDER and RLE ROOT are obtained.
  • step 390 if the INDIRECT bit is set or if the type is a "node" than an error condition is indicated and the lookup terminates as shown in step 400.
  • the lookup is complete and the results stored, as shown in step 410.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Theoretical Computer Science (AREA)
  • Signal Processing (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Radio Relay Systems (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Small-Scale Networks (AREA)

Abstract

A Route Lookup Engine (RLE) for determining a next hop index is disclosed. The RLE receives a lookup key and performs a multi-bit trie search with prefix expansion and capture of a variable stride trie. The data that the RLE returns comprises the next hop information and status flags. The RLE uses a compact, field reusable data structure. The RLE performs both unicast and multicast IP address lookups on Virtual Private Networks. The RLE uses separate indexing and forwarding memories. The upper bound of the search time for the RLE is fixed regardless of the route table size.

Description

    BACKGROUND OF THE INVENTION
  • Network devices such as routers must perform route lookups as part of their processing operations. A route lookup is necessary for every packet received by a router in order to determine where the packet is to be sent next The received packets include a 32-bit wide destination address. The destination address identifies a unique address to which the packet is to be forwarded. A route lookup utilizes the destination address of the received packet in order to identify the next-hop address and exit port in a routing table. Given that the destination address is 32 bits wide, the number of possible destinations is greater than one billion. An IP address comprises a network number and a host number. The network number is of variable length and identifies a unique network. Concatenated to the network number is a host number. The host number identifies a unique host system within a particular network. A received network number part of the address is matched with multiple entries in a routing table to identify the best match. The best match is the one with the largest matching number of bits in the first part in the address. Known methods used to perform this matching are complicated and can take a long time. It would be desirable to provide a route lookup engine which can perform a route lookup quickly and efficiently.
  • BRIEF SUMMARY OF THE INVENTION
  • A Route Lookup Engine (RLE) for determining a next hop index is disclosed. The RLE receives a lookup key and performs a multi-bit trie search with prefix expansion and capture of a variable stride trie. The data that the RLE returns comprises the next hop information and status flags. The RLE uses a compact, field reusable data structure. The RLE performs both unicast and multicast IP address lookups on Virtual Private Networks. The RLE uses separate indexing and forwarding memories. The upper bound of the search time for the RLE is fixed regardless of the route table size.
  • BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING
  • The invention will be more fully understood from the following detailed description taken in conjunction with the accompanying drawings in which:
  • Fig. 1 is a top-level block diagram of an ATM switch;
  • Fig. 2 is a block diagram of a forwarding engine;
  • Fig. 3 is a block diagram of the RLE of the present invention; and
  • Fig. 4 is a flow chart for performing a route lookup according to the present invention.
  • DETAILED DESCRIPTION OF THE INVENTION
  • A Route Lookup Engine (RLE)is disclosed. The RLE may be used in a variety of applications. In a preferred embodiment the RLE is part of an Internet Protocol (IP) routing subsystem for an Asynchronous Transfer Mode (ATM) switch. The routing subsystem of the ATM switch provides a fast and efficient route lookup mechanism. Referring to Fig. 1, a top-level block diagram of a an ATM switch is shown. The ATM switch includes a Input/Output Adapter (IOA) 52. The IOA includes interface optics, Synchronous Optical Network (SONET) framers and High Level Data Link Control (HDLC) controllers. The ATM switch also includes a Input/Output Module (IOM) 10. The IOM includes a plurality of forwarding engines 20. The IOM further includes a timing control unit 30, a DC power conversion and distribution unit 40, a switch fabric interface 50 and control processor 60. The timing control unit 30 of the IOM provides system reference for local timing to the physical layer. The Power Conversion and Distribution unit 40 of the IOM comprises a DC-DC converter. The -48 volts available from the backplane is converted to the required voltages for the module.
  • The switch fabric interface 50 connects each forwarding engine with other cards in the switch via the switch fabric. Data is carried in the form of ATM cells. Switch fabric congestion information is also carried back from the fabric and reformatted/summarized for the forwarding engine.
  • The control processor 60 comprises a Pentium microprocessor as a background general-purpose processor. The forwarding engines 20 will be described in detail below.
  • Referring now to Fig. 2, a block diagram of a single forwarding engine 20 of the IOM is shown. The forwarding engine 20 comprises an inbound Segmentation and Reassembly (SAR) device 150 with associated control memory 160 and packet memory 170, an outbound SAR 180 with associated control memory 190 and packet memory 200, a Packet Header Distributor (PHD) 210, a Packet Classification Engine (PCE) 110, a Route Lookup Engine (RLE) 100, a forwarding processor 130, a cache 140 associated with the forwarding processor and a DRAM/SRAM 120.
  • The forwarding engine 20 processes inbound and outbound packets. For inbound packets, a packet is transferred to the inbound SAR 150, where the packet is stored in the packet memory 170. The inbound SAR 150 sends the highest Quality of Service (QoS) packet to the PHD 210. The forwarding processor 130 checks status of the PHD and determines when inbound packets are ready to be received. The forwarding processor 130 retrieves the highest priority header from PHD FIFOs and stores them in their cache 140. The forwarding processor 130 processes the header and sends required header elements simultaneously to both the RLE 100 and the PCE 110 where they are stored in egress queues. The RLE 100 performs a public or VPN unicast lookup and then a subsequent multicast lookup if required, while the PCE 110 runs filter checks against the header elements.
  • When completed the RLE 100 and PCE 110 pass status to the PHD 210 and store results in an egress queue which is polled by the forwarding processor 130. The operation of the RLE 100 will be described in detail below. The forwarding processor 130 polls and subsequently obtains the route lookup and classification results from which a new header is formulated. This new header is then written to the PHD 210 along with transmit information. The inbound SAR 150 retrieves the newly formatted header, returns it to packet memory 170 with the rest of the packet, and schedules the packet to be segmented and transmitted to the switch fabric.
  • The forwarding engine 20 processes outbound packets as follows. ATM cells are transferred from the switch packet to the outbound SAR 180 where the entire packet is reassembled and stored in packet memory 200. The outbound SAR 180 sends the header of the highest QoS packet to the PHD 210. The forwarding processor 130 checks the status of the PHD 210 and determines outbound packets are ready to be received. The forwarding processor 130 retrieves the highest priority header from the PHD FIFOs and stores them in their cache. The forwarding processor 130 processes the header and may send required header elements simultaneously to both the RLE 100 and the PCE 110 where they are stored in egress queues. The RLE 100 performs a public or VPN unicast lookup and then a subsequent multicast lookup if required, while the PCE 110 runs filter checks against the header elements.
  • When completed the RLE 100 and PCE 110 pass status to the PHD 210, store results in an egress queue which is polled by the forwarding processor 130. The forwarding processor 130 polls and subsequently obtains the route lookup and classification results from which a new header is formulated. This new header is then written to the PHD 210 along with transmit information. The outbound SAR 180 retrieves the newly formatted header, returns it to packet memory 200 with the rest of the packet, and schedules the packet to be transmitted to the physical frame interface.
  • Referring now to Fig. 3, a block diagram of the RLE 100 of the forwarding engine 20 is shown. The RLE is a hardware search engine for determining the next-hop index for the forwarding processor 130. The RLE comprises a forwarding processor (FP) interface 65, input and output FIFOs 40 and 50, a lookup controller 20, a lookup table DRAM 30, and a control processor interface 70. The FP interface 65 is used to load the key in the search engine. The key comprises the address of the root of the tree/the VPN ID unique to the physical port/logical port, the IP destination address, the IP source address, the order of the first level of the tree, and an optional request ID to be returned with the header for sequence error checking. The control processor interface 70 is used for performing table updates between lookups.
  • In operation the FP writes one or more cache lines to a common address for both the RLE and the PCE simultaneously. The FP interface 65 is used to retrieve the search results from both the RLE and PCE simultaneously. They report together from the same address, with a portion of the result provided by the PCE and a portion of the result provided by the RLE.
  • The RLE includes FIFO or other interfaces to enable it to maximize throughput. The input and output FIFOs allow several lookup requests from the FP to be outstanding at a given time. If those packets have different lookup times, the FIFOs smooth the lookup rate, by allowing several header bursts in and out of the RLE. The RLE performs both unicast and multicast IP address lookup on Virtual Private Networks (VPNs). Each VPN is used as a pointer to the root of a table specific to that VPN. Multiple VPNs may map to the same route table via software indirection.
  • To forward IP packets, the Power Processor Controller submits a route lookup request to the PHD. The hardware VST tables of the RLE are used to resolve the request with the resulting indices included with the response from the RLE. Once the indices are made available to the Power Processor Controller, they are used to directly access the forwarding information from the forwarding table.
  • The results from the route lookup comprise the next-hop index, valid lookup/no route found bits, status flags, a multicast lookup preformed bit, statistics information and an optional request ID for sequence error checking.
  • The RLE search is first performed on the IP destination address and subsequently on the IP source address if the packet is a multicast packet. The search utilizes a multi-bit tree search with prefix expansion and capture. The search terminates when a next-hop index is found or the end of the key is reached. Hardware error checking terminates the search and reports an error if an attempt is made to continue past 64 bits of IP header search at the end of the IP source address. An RLE manager of the routing subsystem manages the RLE memory. The RLE memory is used to store the hardware Variable Stride Trie (VST) route tables for each of the configured Virtual Private Networks (VPNs) including the default VPN, VPNO. Each hardware VST is guaranteed a minimum amount of contiguous memory space. The remaining memory space is made available as free pools of memory blocks. As routes are added to and deleted from the VST tables, memory blocks are allocated from and released to these memory pools. The routing subsystem manages the memory to keep fragmentation of the memory at a minimum.
  • The routing subsystem ensures that the forwarding database resident on the memory of the Power Processor Controller remains in synchronization with the hardware VST tables resident on the memory of the RLE. Similarly, the software VST tables resident on the memory of the Kernel processor must remain synchronized with the hardware VST tables of the RLE.
  • The RLE includes Variable Stride Trie (VST) based route lookup tables which reside on a local memory. These VST tables correspond to one specific route. Each entry of the hardware VST table contains an index to the corresponding forwarding entry within a forwarding table resident on the Power Processor Controller memory.
  • For each VPN, a single software VST is maintained in the memory of the kernel processor and a corresponding hardware VST is maintained in the RLE memory. VST route tables are built with a Variable Stride Trie data structure which supports best matching prefixes at a high rate. The VST tables are built by partitioning the search key into discrete bit strings. The search key is the IP address. The number of VST levels specifies the maximum number of memory accesses during the route lookup. Thus, the time to complete the lookup is bounded but not deterministic. The length of each VST level is known as the stride length and determines the number of elements within the node representing that partition.
  • The hardware VST data structures comprise:
    element
    31 30 29 28 24 23 0
    type indirect multicast Order Pointer
    Type: 0 = Leaf, 1= Node
    Order: log2 for nodes, prefix length for leaves
    Pointer: pointer to base address of next node (null if invalid or index to forwarding entry (0xFFFFFF if invalid)
    Figure 00060001
  • The hardware route search data structure comprises nodes, each of which is an array of elements. The number of elements in a node depends on the node order, where order is the number of bits of the search key (destination IP address) used to evaluate this node. Each element has a type indicator that designates the NEXT node as either a leaf or the root of another node. The order indicates the order of the NEXT node if the type of the current node is "node" and the prefix length of the leaf if the type of the current node is "leaf'. The pointer field depends on the element type. If the type is a "leaf', the pointer is an index into the foreground forwarding table and identifies the forwarding entry for the destination address/mask pair. If the type is "node", the pointer is a pointer into search space identifying the address of the next node to be evaluated. The table below lists the meaning of the order and pointer fields for different types.
    Type Order Pointer
    0 (Node) Log2 (# of elements in node at pointer) Points to a node
    1 (Leaf) (Prefix length-1_) of entry pointer Index to forw. entry
  • The RLE table entry comprises:
    Bits 31 30 29 28 24 23 0
    Field P (I) T (1) I (1) Order (5) RLE-Root/Flag.NHP
    where:
  • P:
    Parity bit. Odd parity is optionally written into and read out of the SDRAM by the hardware. {1 bit}
    T:
    Type bit. 0 = node, 1 = leaf. {1 bit}
    I:
    Indirect bit. 1 = indirect, 0 = direct. { 1 bit}
    Order:
    Order of next level corresponding to 1 to 32 {5 bits}
    RLE_ROOT/Flag.NHP:
    There are 2 contexts to this data field depending on the T bit. If a node, it is the Next Node Pointer {24 bits} or if a leaf, it is the RLE flags (8 bits) prepended to the RLE Next Hop Pointer { 16 bits}.
  • The RLE Root is used to start the search. The RLE root contains the root pointer and the order of the first level of the tree. The first lookup address is comprised of the root pointer plus an index The index is derived from the upper bits of the destination address. The number of bits is variable and is indicated by the order of the first level. This address is used to lookup an element from the table. If this element is a leaf then the search is complete and the element is passed back to the PCE/RLE manager. If the element is a node, then the pointer is used for the next lookup.
  • The address for the next lookup is comprised of the above pointer plus an index. The index is derived from the "next" bits from the destination address. The exact bits are derived from the order of the last level, plus the order of this level. Therefore, the order of each level needs to be tracked in order to know the relevant bits of the Destination Address to be used for the index.
  • This address is again used to lookup an element for the table. If this element is a leaf, then the search is complete and the element is passed back to the PCE/RLE manager. If the element is a node, then the pointer is used for the next lookup. This is repeated until a leaf is found. If the number of accesses or levels exceeds a maximum number an error is returned instead.
  • In certain configurations it is desirable to have a method of indirection for the route lookup. This is accomplished by use of the indirect bit. When the indirect bit is set, the leaf flag is ignored, the order field is not used, and the current 24 bit pointer is used as a direct table index, with the route results returned from that indexed location. The indirect function can also be invoked by setting the depth to zero.
  • Referring now to Fig. 4 a flowchart of the presently disclosed method 300 of performing a route lookup is shown. The lookup is started upon receipt of a request at step 310. At step 320 a determination is made as to whether this is the first lookup for this request. If this is not the first lookup for the request than the previous values for the TYPE, INDIRECT, ORDER and RLE ROOT are used, as shown in step 330. If this is the first lookup for this request, then the values for the TYPE, INDIRECT, ORDER and RLE ROOT are taken from the request block data.
  • At step 350 the TYPE bit is evaluated. IF the type is not a "leaf', then steps 420 through 480 et seq. are executed. At step 420 the dram address is set to the RLE ROOT plus zero concatenated with the destination address and the source address.
  • Step 430 is then executed which performs a read from the dram address and new values for the TYPE, INDIRECT, ORDER and RLE ROOT are obtained. At step 440 the current order is set to the current order plus the order.
  • At step 450 a determination is made as to whether the current order is less than or equal to sixty-four. If it is not, then an error condition is indicated and the search terminates. If the current order is less than or equal to sixty-four, then a determination is made.at step 470 as to whether the current order is greater than thirty-two: If not, then another lookup is performed beginning at step 320. If the current order is greater than thirty-two, then step 480 is executed the L bit is set, indicating a multicast operation. After step 480, another lookup is performed beginning at step 320.
  • Referring back to step 350, if the TYPE bit was not a "leaf', then a determination is made at step 360 whether the INDIRECT bit is set. If the INDIRECT bit is not set, then the lookup is complete and the results stored, as shown in step 410.
  • When the indirect bit is set at step 360, step 370 is executed wherein the dram address is set to the RLE ROOT. At step 380 the address of the RLE dram is read and new values for the TYPE, INDIRECT, ORDER and RLE ROOT are obtained. At step 390 if the INDIRECT bit is set or if the type is a "node" than an error condition is indicated and the lookup terminates as shown in step 400. When either the INDIRECT bit is not set or the TYPE bit does not indicate a node, the lookup is complete and the results stored, as shown in step 410.
  • Having described preferred embodiments of the present invention it should be apparent to those of ordinary skill in the art that other embodiments and variations of the presently disclosed embodiment incorporating these concepts may be implemented.
       Accordingly, the invention should not be viewed as limited to the described embodiments but rather should be limited solely by the appended claims.

Claims (21)

  1. A route lookup engine comprising:
    an IP lookup table memory including at least one trie for at least one network;
    a lookup controller in communication with said IP lookup table memory and capable of searching said IP lookup table memory to obtain a next hop index; and
    a forwarding processor interface in electrical communication with said lookup controller, said forwarding processor interface providing a search key to said lookup controller for said lookup controller to begin searching said IP lookup table, said forwarding processor interface receiving said next hop index from said lookup controller.
  2. The route lookup engine of claim 1 wherein the at least one network is a virtual private network.
  3. The route lookup engine of claim 1 further comprising a processor interface in communication with said lookup controller and performing table updates to said IP lookup table memory.
  4. The route lookup engine of claim 1 further comprising a lookup key memory disposed between said forwarding processor interface and said lookup controller.
  5. The route lookup engine of claim 4 wherein the lookup key memory is a FIFO memory.
  6. The route lookup engine of claim 1 further comprising a next hop result memory disposed between said lookup controller and said forwarding processor interface.
  7. The route lookup engine of claim 6 wherein the next hop result memory is a FIFO memory.
  8. The route lookup engine of claim 1 further comprising a control/status register in communication with said processor interface, said forwarding processor interface and said lookup controller.
  9. The route lookup engine of claim 1 wherein said lookup controller performs a unicast lookup.
  10. The route lookup engine of claim 1 wherein said lookup controller performs a multicast lookup.
  11. The route lookup engine of claim 1 wherein said search key comprises an IP address.
  12. The route lookup engine of claim 1 wherein said lookup memory comprises a Dram.
  13. The route lookup engine of claim 1 wherein said at least one trie comprises a variable stride trie.
  14. The route lookup engine of claim 1 wherein said lookup controller performs prefix expansion and capture when searching said IP lookup table memory.
  15. A method for operating a route lookup engine comprising the steps of:
    receiving a route lookup request;
    performing a route lookup comprising the steps of:
    determining if the lookup is the first one for said request;
    using current block data for said request when the lookup is the first lookup for said request, using prior block data when the lookup is not the first lookup for said request;
    determining if said block data indicates a leaf, and when said block data does not indicate a leaf performing the steps of:
    providing a route table lookup address comprising a route lookup engine root from said block data plus an address comprising a zero bit concatenated with a destination address concatenated with a source address minus the current order data from said block data;
    performing a read from said lookup table, said read providing new block data;
    setting the current order data to the present current order data plus the order data from the new block data;
    returning to said step of determining if the lookup is the first one for said request;
    performing the following steps when said block data does indicate a leaf:
    determining if the indirect bit is set in said block data and when said indirect bit is set performing the steps of:
    providing a route table lookup address comprising a route table lookup engine route from said block data;
    performing a read from said lookup table, said read providing new block data;
    storing results if said indirect bit does not equal a one and the type bit does not equal node; and
    performing the step of storing results when said indirect bit is not set.
  16. The method of claim 15 wherein the step of setting the current order data further comprises the step of indicating an error condition when said current order data is less than or equal to sixty-four.
  17. The method of claim 15 wherein the step of setting the current order data further comprises the step of setting a multicast bit when said current order is greater than thirty-two.
  18. The method of claim 15 wherein the step of storing results if said indirect bit does not equal a one and the type bit does not equal node further comprises the step of indicating an error condition if said indirect bit equals a one or said type bit equals a node.
  19. The method of claim 15 wherein said step of performing a route lookup comprises performing a unicast lookup.
  20. The method of claim 15 wherein said step of performing a route lookup comprises performing a multicast lookup.
  21. A method for operating a route lookup engine comprising the steps of
    receiving a route lookup request;
    performing a route lookup comprising the steps of
    determining if the lookup is the first one for said request;
    using current block data for said request when the lookup is the first lookup for said request, using prior block data when the lookup is not the first lookup for said request;
    determining if said block data indicates a leaf, and when said block data does indicate a leaf performing the steps of:
       determining if an indirect bit is set in said block data and when said indirect bit is set performing the steps of:
    providing a route table lookup address comprising a route table lookup engine route from said block data;
    performing a read from said lookup table, said read providing new block data;
    storing results if said indirect bit does not equal a one and the type bit does not equal node; and
    performing the step of storing results when said indirect bit is not set.
EP00310758A 1999-12-13 2000-12-04 Route lookup engine Ceased EP1122927A3 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US459441 1999-12-13
US09/459,441 US6711153B1 (en) 1999-12-13 1999-12-13 Route lookup engine

Publications (2)

Publication Number Publication Date
EP1122927A2 true EP1122927A2 (en) 2001-08-08
EP1122927A3 EP1122927A3 (en) 2004-04-14

Family

ID=23824784

Family Applications (1)

Application Number Title Priority Date Filing Date
EP00310758A Ceased EP1122927A3 (en) 1999-12-13 2000-12-04 Route lookup engine

Country Status (4)

Country Link
US (1) US6711153B1 (en)
EP (1) EP1122927A3 (en)
JP (2) JP3926558B2 (en)
CA (1) CA2326928C (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003079618A2 (en) * 2002-03-15 2003-09-25 Globespan Virata Incorporated System and method for longest prefix match internet protocol lookup
EP1453253A1 (en) * 2003-02-28 2004-09-01 Samsung Electronics Co., Ltd. Packet classification apparatus and method using field level tries
US7493412B2 (en) 2002-09-12 2009-02-17 International Business Machines Corporation Method for processing a data packet

Families Citing this family (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7000034B2 (en) * 2000-03-02 2006-02-14 Agere Systems Inc. Function interface system and method of processing issued functions between co-processors
US6850516B2 (en) * 2000-03-02 2005-02-01 Agere Systems Inc. Virtual reassembly system and method of operation thereof
DE10011667C2 (en) * 2000-03-10 2002-11-21 Infineon Technologies Ag High Speed Router
US6798777B1 (en) * 2000-04-17 2004-09-28 Juniper Networks, Inc. Filtering and route lookup in a switching device
US6957272B2 (en) * 2000-05-24 2005-10-18 Alcatel Internetworking (Pe), Inc. Stackable lookup engines
US6555511B2 (en) * 2000-06-19 2003-04-29 Lance L. Renfrow Stable hydrotropic surfactants comprising alkylamino propionate
US6880064B1 (en) * 2000-06-21 2005-04-12 Mosaid Technologies, Inc. Method and apparatus for physical width expansion of a longest prefix match lookup table
JP4225681B2 (en) * 2000-12-06 2009-02-18 富士通株式会社 Virtual closed network construction method and apparatus, and relay apparatus
US6940854B1 (en) * 2001-03-23 2005-09-06 Advanced Micro Devices, Inc. Systems and methods for determining priority based on previous priority determinations
US7389360B1 (en) 2001-11-05 2008-06-17 Juniper Networks, Inc. Context switched route lookup key engine
US7356033B2 (en) * 2002-11-21 2008-04-08 Lucent Technologies Inc. Method and apparatus for performing network routing with use of power efficient TCAM-based forwarding engine architectures
US7782853B2 (en) * 2002-12-06 2010-08-24 Stmicroelectronics, Inc. Apparatus and method of using fully configurable memory, multi-stage pipeline logic and an embedded processor to implement multi-bit trie algorithmic network search engine
US7924839B2 (en) * 2002-12-06 2011-04-12 Stmicroelectronics, Inc. Mechanism to reduce lookup latency in a pipelined hardware implementation of a trie-based IP lookup algorithm
US7620040B2 (en) * 2002-12-11 2009-11-17 Aspen Networks, Inc. Application non disruptive task migration in a network edge switch
DE10260604B4 (en) * 2002-12-23 2006-09-21 Infineon Technologies Ag Multi-channel processor
US7702882B2 (en) * 2003-09-10 2010-04-20 Samsung Electronics Co., Ltd. Apparatus and method for performing high-speed lookups in a routing table
US7586911B2 (en) * 2003-10-17 2009-09-08 Rmi Corporation Method and apparatus for packet transmit queue control
GB0524845D0 (en) * 2005-12-06 2006-01-11 Univ Belfast Sorting apparatus and method
US7644343B2 (en) * 2006-01-17 2010-01-05 Rajugopal Gubbi Error resilience methods for multi-protocol encapsulation forward error correction implementations
CN1946052A (en) * 2006-10-19 2007-04-11 北京四达时代软件技术有限公司 One-way network video frequency video on demand portal system
TR201209797T2 (en) * 2009-12-22 2013-01-21 Fehm� Bazlama�Ci C�Neyt Systolic array architecture for fast IP lookup.
US10135734B1 (en) 2015-12-28 2018-11-20 Amazon Technologies, Inc. Pipelined evaluations for algorithmic forwarding route lookup

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0551243A2 (en) * 1992-01-10 1993-07-14 Digital Equipment Corporation Address recognition engine

Family Cites Families (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5842224A (en) 1989-06-16 1998-11-24 Fenner; Peter R. Method and apparatus for source filtering data packets between networks of differing media
AU620994B2 (en) 1989-07-12 1992-02-27 Digital Equipment Corporation Compressed prefix matching database searching
GB9326476D0 (en) * 1993-12-24 1994-02-23 Newbridge Networks Corp Network
US5412654A (en) 1994-01-10 1995-05-02 International Business Machines Corporation Highly dynamic destination-sequenced destination vector routing for mobile computers
US5509123A (en) * 1994-03-22 1996-04-16 Cabletron Systems, Inc. Distributed autonomous object architectures for network layer routing
US5787430A (en) 1994-06-30 1998-07-28 International Business Machines Corporation Variable length data sequence backtracking a trie structure
US5870739A (en) 1996-09-20 1999-02-09 Novell, Inc. Hybrid query apparatus and method
US5905725A (en) * 1996-12-16 1999-05-18 Juniper Networks High speed switching device
US6011795A (en) 1997-03-20 2000-01-04 Washington University Method and apparatus for fast hierarchical address lookup using controlled expansion of prefixes
US5946679A (en) 1997-07-31 1999-08-31 Torrent Networking Technologies, Corp. System and method for locating a route in a route table using hashing and compressed radix tree searching
US6266706B1 (en) * 1997-09-15 2001-07-24 Effnet Group Ab Fast routing lookup system using complete prefix tree, bit vector, and pointers in a routing table for determining where to route IP datagrams
JP3186681B2 (en) * 1997-12-25 2001-07-11 日本電気株式会社 Route search circuit and communication control device
US6061712A (en) 1998-01-07 2000-05-09 Lucent Technologies, Inc. Method for IP routing table look-up
US6052683A (en) 1998-02-24 2000-04-18 Nortel Networks Corporation Address lookup in packet data communication networks
US6192051B1 (en) * 1999-02-26 2001-02-20 Redstone Communications, Inc. Network router search engine using compressed tree forwarding table
US6463067B1 (en) * 1999-12-13 2002-10-08 Ascend Communications, Inc. Submission and response architecture for route lookup and packet classification requests

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0551243A2 (en) * 1992-01-10 1993-07-14 Digital Equipment Corporation Address recognition engine

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003079618A2 (en) * 2002-03-15 2003-09-25 Globespan Virata Incorporated System and method for longest prefix match internet protocol lookup
WO2003079618A3 (en) * 2002-03-15 2004-09-10 Globespan Virata Inc System and method for longest prefix match internet protocol lookup
US7493412B2 (en) 2002-09-12 2009-02-17 International Business Machines Corporation Method for processing a data packet
EP1453253A1 (en) * 2003-02-28 2004-09-01 Samsung Electronics Co., Ltd. Packet classification apparatus and method using field level tries
CN100385880C (en) * 2003-02-28 2008-04-30 三星电子株式会社 Packet classification apparatus and method using field level tries
US7415020B2 (en) 2003-02-28 2008-08-19 Samsung Electronics Co., Ltd. Packet classification apparatus and method using field level tries

Also Published As

Publication number Publication date
CA2326928A1 (en) 2001-06-13
CA2326928C (en) 2005-01-11
JP4542539B2 (en) 2010-09-15
US6711153B1 (en) 2004-03-23
JP2001223750A (en) 2001-08-17
JP2007110745A (en) 2007-04-26
JP3926558B2 (en) 2007-06-06
EP1122927A3 (en) 2004-04-14

Similar Documents

Publication Publication Date Title
US6711153B1 (en) Route lookup engine
JP3640299B2 (en) A proposal and response architecture for route lookup and packet classification requests
US6650642B1 (en) Network relaying apparatus and network relaying method capable of high-speed routing and packet transfer
US6633565B1 (en) Apparatus for and method of flow switching in a data communications network
US6683885B1 (en) Network relaying apparatus and network relaying method
JP3734704B2 (en) Packet classification engine
US5978951A (en) High speed cache management unit for use in a bridge/router
US6192051B1 (en) Network router search engine using compressed tree forwarding table
CA2274962C (en) High speed variable length best match look-up in a switching device
US5909440A (en) High speed variable length best match look-up in a switching device
US20050171937A1 (en) Memory efficient hashing algorithm
JP3340846B2 (en) ATM-LAN, server and ATM address management method
US7236492B2 (en) Configurable packet processor
JP3371006B2 (en) Table search method and router device
US8265072B2 (en) Frame switching device
US6658003B1 (en) Network relaying apparatus and network relaying method capable of high-speed flow detection
JP2001045061A (en) Communication node device
JPH0685819A (en) Computer system
JP2000032004A (en) Node device
US6671277B1 (en) Network relaying apparatus and network relaying method capable of high quality transfer of packets under stable service quality control
Tantawy et al. On the design of a multigigabit IP router
JP2000312225A (en) Packet repeater
US7613200B1 (en) Method and apparatus using a random indication to map items to paths and to recirculate or delay the sending of a particular item when a destination over its mapped path is unreachable
JPH03235444A (en) Address filter device
US20040143592A1 (en) Method for processing redundant packets in computer network equipment

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A2

Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE TR

AX Request for extension of the european patent

Free format text: AL;LT;LV;MK;RO;SI

PUAL Search report despatched

Free format text: ORIGINAL CODE: 0009013

AK Designated contracting states

Kind code of ref document: A3

Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE TR

AX Request for extension of the european patent

Extension state: AL LT LV MK RO SI

17P Request for examination filed

Effective date: 20040927

17Q First examination report despatched

Effective date: 20041025

AKX Designation fees paid

Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE TR

APBN Date of receipt of notice of appeal recorded

Free format text: ORIGINAL CODE: EPIDOSNNOA2E

APBR Date of receipt of statement of grounds of appeal recorded

Free format text: ORIGINAL CODE: EPIDOSNNOA3E

APAF Appeal reference modified

Free format text: ORIGINAL CODE: EPIDOSCREFNE

APAF Appeal reference modified

Free format text: ORIGINAL CODE: EPIDOSCREFNE

APBT Appeal procedure closed

Free format text: ORIGINAL CODE: EPIDOSNNOA9E

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION HAS BEEN REFUSED

18R Application refused

Effective date: 20100303